Blind source separation by sensor-signal identity mapping by auto-encoder with hidden-layer pruning

Author

Yasui, Syozo

Author_Institution

Graduate Sch. of Life Sci. & Syst. Eng., Kyushu Inst. of Technol., Iizuka, Japan

Volume

2

fYear

2002

fDate

6/24/1905 12:00:00 AM

Firstpage

1305

Lastpage

1309

Abstract

A new non-information-theoretic approach is described for the blind source separation (BSS) problem. It is based on an auto-encoder neural network which incorporates a pruning algorithm. Hidden units are nonlinear, and ones that survive the pruning become the source extractors. As such, no assumption is needed for the number of sources. Simulation results show that the auto-encoder can make BSS for a broad class of source-signal mixtures without changing the nonlinear activation function of the hidden units

Keywords

deconvolution; encoding; feedforward neural nets; multilayer perceptrons; neural net architecture; signal sources; transfer functions; auto-associative neural network; auto-encoder neural network; blind source separation; hidden-layer pruning algorithm; noninformation-theoretic approach; nonlinear activation function; nonlinear hidden units; sensor-signal identity mapping; simulation; source extractors; source-signal mixtures; Blind source separation; Data compression; Decoding; Independent component analysis; Maximum likelihood estimation; Modeling; Neural networks; Principal component analysis; Source separation; Systems engineering and theory;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 2002. IJCNN '02. Proceedings of the 2002 International Joint Conference on

Conference_Location

Honolulu, HI

ISSN

1098-7576

Print_ISBN

0-7803-7278-6

Type

conf

DOI

10.1109/IJCNN.2002.1007683

Filename

1007683